concurring and dissenting:

I

At the outset I should say that I am able to concur in the result of this case only because it was heard in two separate and distinct parts, and amounts to two cases heard by en banc courts of different composition. The procedural aspect of the case with respect to en banc determination was heard by the judges not disqualified in reg*908ular active service, while the case on the merits of the appeal was heard by them and, in addition, Judge Bryan, a Senior Judge. Had that part of the case with respect to en banc determination been determined the way I think it should have been, then we would not have reached the merits for en banc determination would have been denied for failure of a majority.

As to the matters in the opinion of the majority with respect to en banc determination, I respectfully dissent for reasons which I will set forth below.

The fact that I concur in the grant of a new trial to Eastern does not mean that I associate myself with all of the opinion of the majority.

I concur in that part of the opinion of the majority reversing for a new trial on account of consolidation of the cases as well as the result obtained by the majority. If the subject were properly preserved, I do not disagree with the majority opinion as it adopts the reasoning of the dissenting panel opinion with respect to the conduct of the trial and especially the statements of counsel during oral argument and perhaps elsewhere. As Judge Phillips has correctly set forth in his dissenting opinion, however, there were no sufficient contemporaneous objections made to the actions of the attorneys. Neither were there sufficient, if any, contemporaneous motions for mistrial. The absence of contemporaneous objections and motions for mistrial I think precludes consideration on appeal. It simply deprives the other side of a fair trial for an attorney to sit through an inflammatory argument which could have been and should have been stopped instantly, and, then, after the argument is over, make objection to it. At the very least, such lack of action deprives the trial court of an opportunity to see that the trial is conducted in a calm and fair manner. I can think of no other aspect which more demands the enforcement of the contemporaneous objection rule and the rule requiring motions for mistrial than does argument of counsel. Not to enforce that rule simply opens the door for turning trials into shouting matches, and I would at once close it in this en banc decision. If that were the only subject at hand, I would affirm the judgment of the district court because of lack of contemporaneous objections and motions for mistrial.

II

A.

I do not subscribe to the majority’s theory that the Western Pacific Case has contributed to our uncertainty.

B.

Congress, pursuant to its constitutional power to establish inferior courts, has provided that the Courts of Appeals shall hear cases “by a court or panel of not more than three judges, unless a ... rehearing before the court in banc is ordered by a majority of the circuit judges of the circuit who are in regular active service.” 28 U.S.C. § 46(c). After this case was decided by a panel of three judges, the appellee, Eastern Airlines, petitioned for rehearing and suggested rehearing en banc. A poll of the ten circuit judges of this circuit in regular active service yielded five votes to grant rehearing, four votes to deny rehearing, and one disqualification. Because five votes does not constitute a majority of ten, the number of circuit judges in regular active service at the time of the poll, I respectfully dissent from the court’s decision to rehear this case en banc.

In response to various problems concerning the proper role of retired judges in the affairs of the Courts of Appeal, Congress enacted a bill, proposed by the Judicial Conference of the United States, the purpose of which was to clarify the status of circuit and district judges who retired from regular active service. Act of Nov. 13, 1963, Pub.L. 88-176, 77 Stat. 331. S.Rep. No. 596, 87th Cong., 1st Sess., reprinted in 1963 U.S.Code Cong, and Ad.News, 1105, 1106. This legislation defined the composition of each of the various Courts of Appeal as consisting of the circuit judges of the circuit in regular active service, that is, those judges not retired. Pub.L. 88-176, § 1(a) *909(codified at 28 U.S.C. § 43(b)). The Act then amended 28 U.S.C. § 46(c) to put control of the en banc proceedings in the hands of “a majority of the circuit judges of the circuit who are in regular active service.” Id. at § 1(b). All the amendment did, as it affects us here, was to insert “regular” between “in” and “active” in the then existing statute.

Prior to the enactment of that amendment, the Supreme Court had discussed en banc courts in Textile Mills Corp. v. Commissioner, 314 U.S. 326, 62 S.Ct. 272, 86 L.Ed. 249 (1941), and in the Western Pacific Railroad Case, 345 U.S. 247, 73 S.Ct. 656, 97 L.Ed. 986 (1953), as well as in Shenker hereinafter mentioned. Textile Mills held that the statutes with respect to the composition of the Courts of Appeals did not prohibit a Court of Appeals of more than three judges from hearing a case en banc, and the Western Pacific case construed the predecessor of the statute involved here as simply a grant of power to order hearings and rehearings en banc and to establish the procedure governing the exercise of that power. 345 U.S. at 267, 73 S.Ct. at 666. Its holding was simply that a Court of Appeals could not restrict the initiation of the en banc procedure to the court itself, and that a decision striking out an application by a party for en banc consideration was error. In discussion the Court did construe § 46(c) as neither forbidding nor requiring each active judge of a Court of Appeals to entertain each petition for en banc consideration, which is entirely consistent with its later ruling in Shenker.

In Shenker v. Baltimore & Ohio RR Co., 374 U.S. 1, 4 — 5, 83 S.Ct. 1667,1670-1671,10 L.Ed.2d 709 (1962), the Supreme Court affirmed a decision of the Third Circuit which required an affirmative vote of an absolute majority of the active members of the court and which did not require each member of the court to vote on petitions for hearing en banc. At the time Shenker filed his petition for rehearing en banc, eight judges were in active service. Four of the eight voted to rehear the case; two voted not to rehear; and two abstained for reasons that do not appear either in the Supreme Court’s or the Court of Appeals’ opinion. See Shenker v. Baltimore & O.R. Co., 303 F.2d 596, 600 (1962). On this vote, the Third Circuit denied the petition for rehearing en banc, and the Supreme Court found no error in the denial. 374 U.S. at 4-5, 83 S.Ct. at 1670-1671. The Supreme Court also found no error in the Third Circuit’s procedure for polling its judges, which did not require a judge to vote on a petition. Id. at 5, 83 S.Ct. at 1670. The Court found such a procedure to be within the discretion of the Court of Appeals.

I do not construe Shenker to hold that a court may by rule set up the requirement of an absolute majority of the circuit judges in regular active service and then to have founded its decision upon the rule. In the first place, so far as I can ascertain, there was no rule in the Third Circuit except a near literal copying of the words of the statute.1 It was merely a practice of that circuit which was given effect by decision in the Shenker case. Thus, the decision of the Third Circuit was that an absolute majority was required, and this was affirmed by the Supreme Court. Both the holding of the Third Circuit and the Supreme Court are entirely consistent with the wording of the statute. Second, even discounting the two Shenker abstentions, the vote at the very best for the petitioner was four for and two against if we count the two abstentions as votes against granting the petition. It is at once apparent that four to two is a majority of those judges not disqualified, and so the fact situation presented to the Supreme Court in Shenker is the same as presented to us here. Since the Shenker case offered a perfect opportunity for the Supreme Court to adopt the view the majority now does, and it did not do so, I suggest that, because the adoption of the majority view here would have changed the result of the Shenker case, the Supreme Court has re*910jected the position the majority now embraces.

The Second Circuit decided a similar question in Zahn v. International Paper Co., 469 F.2d 1033, 1040 (2d Cir. 1972), aff’d on other grounds, 414 U.S. 291, 94 S.Ct. 505, 38 L.Ed.2d 511 (1973), in a way that supports my contention that the motion for rehearing in this case failed for want of a majority. When the petition for rehearing was considered in the Second Circuit, the court consisted of nine judges. At the time, there was one vacancy.2 One judge did not vote because of disqualification. Of the seven judges who voted, four voted to rehear the case, and three voted not to rehear. Despite this four-to-three vote in favor of rehearing en banc, the Second Circuit denied the petition “for want of an affirmative vote ‘by a majority of the circuit judges of the circuit who are in regular active service.’ ” Zahn, supra at 1040, quoting 28 U.S.C. § 46(c). The Second Circuit has adhered to its literal reading of § 46(c). E. g., Boyd v. Lefrak Organization, 517 F.2d 918 (1975) (rehearing en banc denied; four affirmative votes to rehear, three negative votes, one disqualification, one vacancy).

Besides Zahn, a majority of circuits which have confronted the problem now before us have read § 469(c) as requiring an absolute majority of the circuit judges in regular active service to order a rehearing en banc. Copper and Brass Fabricators Council, Inc. v. Department of Treasury, 679 F.2d 951 (D.C.Cir. 1982), reh’g denied, unpublished order, No. 81-2091 (Aug. 3, 1982) (five votes to rehear; three votes not to rehear; two judges not participating in the matter); Clark v. American Broadcasting Co, 684 F. 2d 1208, 1226 (6th Cir. 1982) (rehearing denied; five votes in favor of rehearing, four against, and one judge disqualified); Curtiss-Wright Corp. v. General Elec. Co., 599 F.2d 1259 (3d Cir. 1979) (rehearing denied; four votes to rehear en banc, three votes not to rehear, two judges not participating); Porter County Chapter of Izaak Walton League v. Atomic Energy Comm’n, 515 F.2d 513, 533-34 (7th Cir. 1975) (rehearing denied; four votes for rehearing en banc, three votes against, one judge taking no part), rev’d on other grounds, sub nom. Northern Indiana Public Service Co. v. Porter County Chapter of Izaak Walton League, 423 U.S. 12, 96 S.Ct. 172, 46 L.Ed.2d 156 (1975). See also Internal Operating Procedures for the United States Court of Appeals for the Third Circuit, ch. IX, B.4.b. (rev. Sept. 4, 1980) (effect of non-response to a poll of circuit judges on decision whether to rehear a case en banc is the same as a vote not to rehear). Accord Maris, Hearing and Rehearing In Banc, 14 F.R.D. 91, 93 (1954) (the Third Circuit follows § 46(c) literally; hearing or rehearing en banc is ordered only with the concurrence of an absolute majority of the circuit judges in regular active service).

A minority of the circuits that have dealt with the question have authorized en banc review upon less than an absolute majority of the circuit judges in regular active service. In the Eighth Circuit, “a majority of the judges ... in regular active service who are actively participating in the affairs of the court and who are not disqualified in the particular case or controversy may order a ... rehearing en banc...." Rule 16(a) of the U.S. Court of Appeals for the Eighth Circuit (1981).

Although the Ninth Circuit apparently does not reveal the results of a poll of its members on whether to hear or rehear a case en banc, one of its members has described the circuit’s policy in a footnote to a dissent from the denial of rehearing en banc. Ford Motor Co. v. Federal Trade Comm’n, 673 F.2d 1008, 1012, n. 1 (9th Cir. 1982) (Reinhardt, J., dissenting). The Ninth Circuit rehears cases en banc only when a majority of its active members who are not disqualified vote to do so.

The First Circuit, in United States v. Martorano, 620 F.2d 912 (1980), granted rehearing en banc when the vote was two for, one against, and one vacancy. Thus, *911that court has joined neither the majority nor minority.

Strong practical considerations also support the plain reading of the majority requirement of § 46(c), and of Fed.R.App.P. 35(a), which was adopted pursuant to the statute. “The principal utility of determinations by the courts of appeals en banc is to enable the court to maintain its integrity as an institution by making it possible for a majority of its judges always to control and thereby to secure uniformity and continuity in its decisions.... ” United States v. American-Foreign S.S. Corp., 363 U.S. 685, 689-90, 80 S.Ct. 1336, 1339-1340, 4 L.Ed.2d 1491 (1960) (quoting from Maris, Hearing and Rehearing In Banc, 14 F.R.D. 91, 96 (1954). In this connection, since the very direction of the court is in the hands of itself assembled en banc, less than a majority does not hold so steady a course.

Section 46(c) allows a circuit to achieve uniformity within itself on questions of importance by acting as a full court rather than as a three-judge panel. Zahn, supra, at 1041 (Mansfield, J.). If fewer than a majority of its judges in regular active service may order en banc determination, as is now the case with us, the en banc procedure will not so surely achieve a uniform body of law, relatively immune from shifting changes in membership or disqualifications in the court. Furthermore, an important control on the invocation of the en banc process, so costly in time and resources, will be lost. Given that en banc hearing is an extraordinary measure, see F.R.A.P. 35(a), the requirement of an absolute majority of all the judges in regular active service would help limit en banc review to those cases most meriting exceptional treatment.

In 1973, the Judicial Conference of the United States approved a proposed bill to amend § 46(c) to “make clear that a majority of the judges in regular active service who are entitled to vote should be sufficient to en banc a case.” 1973 Reports of the Proceedings of the Judicial Conference of the United States 47. Although by advocating such legislation it expressed its approval of the rule that this circuit adopts today, the Judicial Conference acknowledged that Congress was the competent lawmaker in this area, and, further, that the law was to the contrary, otherwise no legislation would have been necessary. The Director of the Administrative Office of the United States Courts transmitted a draft of the proposed legislation to both houses of the 93d Congress, which in turn referred the draft to the respective Committees on the Judiciary. 119 Cong.Rec. 32,129 (House of Representatives) (Oct. 1, 1973); 119 Cong.Rec. 32,153 (Senate) (Oct. 1, 1973). The draft legislation was introduced as H.R. 10805, 93d Cong., 1st Sess., 119 Cong. Rec. 33,430 (Oct. 9, 1973), and was referred to the Subcommittee on Courts, Civil Liberties, and the Administration of Justice, which took no action on the bill other than to request an opinion from the Department of Justice. The bill thus died in the 93d Congress.

Since 1973, two Congresses have seen fit to amend § 46(c), without addressing themselves to the specific problem posed by the Judicial Conference’s draft legislation. See The Federal Courts Improvement Act of 1982, Pub.L. 97-164, § 205, 96 Stat. 25, 53 (1982); Act of Oct. 20, 1978, Pub.L. 95-486, § 5, 92 Stat. 1629 (1978). If any inference is to be drawn from the subsequent legislative history of § 46(c), it is that Congress, the only body competent to change the statute, was aware of the construction of the statute sought to be changed by the Judicial Conference and chose not to change it.

Despite the plain language of the statute and its literal interpretation by a great majority of the circuits, the majority insists that the en banc court comprises the judges in regular active service other than those judges who are disqualified. It would so read the statute because to require an absolute majority would have the effect of treating a disqualification or an abstention as though it were a no vote. Mindful of a judge’s duty to disqualify himself whenever his impartiality might reasonably be questioned, 28 U.S.C. § 455, the majority takes the position that to treat a disqualification as a no vote would allow a judge improperly *912to affect the determination of the cause from which he is barred. At 904.

I think the majority’s concern is legitimate but misdirected.

The clear import of 28 U.S.C. § 455 is to prohibit a judge’s participation in the determination of a case within the purview of that statute. Although a disqualification may be viewed as a no vote, that effect is entirely incidental. We must assume that a judge will recuse himself whenever required to do so, and when he is disqualified, the incidental effect is always one way, namely, to leave the absolute majority requirement in effect with one less possible vote to rehear.

But the majority’s reading of the statute does not cure the mischief it assumes. Under the majority decision, five votes to rehear are needed in this case only because one judge disqualified himself. If that one judge had not disqualified himself, then six votes would have been needed to rehear the case. The result of that one judge’s disqualification was to reduce the number of votes needed to rehear the case en banc from six to five. Viewed from the other side of the case, then, that judge’s disqualification had the immediate effect, albeit unintended and incidental, of a yes vote. The disqualification made en banc review easier to obtain by one vote, and according to the majority’s logic, it is as though the disqualified judge voted yes and is affecting the outcome of the cause from which he is barred, thereby contravening 28 U.S.C. § 455.3 Thus, I suggest the majority’s reasoning, that its construction of § 46(c) avoids a conflict with § 455, is patently self defeating, and reveals the correct conclusion which is that § 455 has nothing to do with how this case should be decided.

I thus respectfully dissent to the grant of en banc review.

By way of summary the Court states its conclusions on the issues considered at the en banc rehearing which took place on February 8, 1983:

I. The vote of five members for, four against, and one member disqualified, and hence not voting, which had previously taken place, constituted a determination by a majority of the circuit judges who are in regular active service ordering rehearing en banc.

On that issue, the vote was six (Chief Judge Winter, Judge Russell, Judge Hall, Judge Murnaghan, Judge Sprouse, and Judge Chapman) for and two (Judge Widener and Judge Phillips) against.

2. The judgments of the district court in the Arnold and Mihalek1 personal injury actions against Eastern are reversed and remanded for retrial on the issue of the proper amounts of compensatory damages.

On that issue, the vote was five (Chief Judge Winter, Judge Russell, Judge Widener, Judge Murnaghan and Judge Chapman) for and four (Judge Hall, Judge Phillips, Judge Sprouse and Judge Bryan) against.

It should be noted that Judge Ervin was disqualified and therefore did not participate in any of the proceedings in the case. Furthermore, Judge Bryan, as a Senior Circuit Judge who sat as a member of the three-judge panel, took part in the consideration as to issue 2, pursuant to 28 U.S.C. § 46(c). Judge Bryan, there being no statutory authorization for him to do so, did not take part in the consideration as to issue 1.

The several opinions of Judges Widener, Hall, Phillips and Murnaghan which follow are devoted to the various rationales which prompted the votes of the several members of the Court.

In the United States District Court for the Western District of North Carolina, there were consolidated for trial several actions arising as a consequence of an Eastern Air Lines airplane crash at Charlotte, North Carolina on September 11, 1974:

1) Three personal injury actions against Eastern.

2) An action for indemnification or contribution by insurance carriers of Eastern against the United States and four air traffic controllers.

3) Third party claims against the United States for contribution filed by Eastern in two of the accident cases.

*902The panel of the Fourth Circuit which heard the appeal affirmed judgments on behalf of the plaintiffs in two of the accident cases, reversed and remanded the third for a new trial on the issue of compensatory damages and affirmed judgments in favor of the United States and the air traffic controllers. Arnold v. Eastern Airlines, Inc., 681 F.2d 186 (4th Cir. 1982). There was a dissent at the panel level limited to the affirmance of awards for plaintiffs in the two accident cases. Id. at 206.

The case is now before the Court as a consequence of Eastern’s request for rehearing en banc, contesting the judgments for the two accident case plaintiffs.

As a matter preliminary to consideration of the issues on the merits, we have to decide whether the case has in fact achieved the status of one entitled to en banc rehearing. The Federal Rules of Appellate Procedure in Rule 35 provide that a “majority of the circuit judges who are in regular active service may order that an appeal ... be ... reheard by the court of appeals in banc.”

At the time of voting on the suggestion for en banc rehearing there were ten individuals occupying positions with the court satisfying description as circuit judges in regular active service. When the poll was taken, one court member, Judge Ervin, recused himself, did not vote, and has since that time remained disqualified. Of the nine remaining circuit judges in regular active service five (Chief Judge Winter, Judge Russell, Judge Widener, Judge Murnaghan and Judge Chapman) voted in favor of rehearing en banc; four circuit judges in regular active service (Judge Butzner, Judge Hall, Judge Phillips and Judge Sprouse) voted to deny rehearing en banc.

The court tentatively determined that the case had achieved en banc rehearing status, and now, following consideration of arguments from counsel, confirms that tentative determination. Joining the author of this opinion in that conclusion are Chief Judge Winter, Judge Russell, Judge Hall, Judge Sprouse and Judge Chapman. Judge Widener and Judge Phillips are of the contrary view and would hold that the vote on the suggestion of an en bane rehearing failed for want of a majority of the circuit judges in regular active service. Five, Judge Widener and Judge Phillips contend, do not constitute a majority of ten.1

At the outset, it should be observed that the majority’s decision has been taken for the purposes of the present case. It is, of course, entitled to the precedential weight attaching to any decision of the court. However, in Shenker v. Baltimore & Ohio R. Co., 374 U.S. 1, 4-5, 83 S.Ct. 1667, 1670-1671, 10 L.Ed.2d 709 (1963), there appears a suggestion that the area may well be one in which achieving fulfillment of our administrative responsibilities would allow us by rule to select, as a quorum for purposes of ascertaining a majority, when votes on suggestions for hearings or rehearings en banc are taken, either (a) all judges in regular active service, including those disqualified for the purposes of the particular case or (b) all judges otherwise in regular active service who are not, for the purposes of the particular case, disqualified from participating in any way.2 Accordingly, we do not, by our decision today preclude a possible change in practice, brought about by adoption of a rule of general ap*903plicability. Nor, of course, do we have occasion to determine whether applicable statutory language would permit or forbid such a change in practice. The uncertainties are not minimized by the Supreme Court’s observation in Western Pacific Rr. Corp. v. Western Pacific Rr. Co., 345 U.S. 247, 260, 73 S.Ct. 656, 662, 97 L.Ed. 986 (1953):

The en banc power ... is ... a necessary and useful power — indeed too useful that we should ever permit a court to ignore the possibilities of its use in cases where its use might be appropriate.

We now have no general rule, and must decide without the benefit of such a generally promulgated guide to practice,3 the proper result in the case sub judice.

We find the answer primarily in the language of the applicable statute. 28 U.S.C. § 46(c):

Cases and controversies shall be heard and determined by a court or panel of not more than three judges, unless a hearing or rehearing before the court in banc is ordered by a majority of the circuit judges of the circuit who are in regular active service. A court in banc shall consist of all circuit judges in regular active service.... 4

Concentrating on the second sentence of § 46(c), we perceive insoluble difficulties in any other reading of the statute, at least in its present posture, without any contrary indication from a rule or regulation, than that there shall be excluded, for quorum ascertainment purposes, any disqualified judge when a vote on a suggestion for hearing or rehearing en banc takes place. Judge Ervin has not sat when the case was reheard en banc. No responsible suggestion could be made that he, who was disqualified and consequently took no part *904whatever in the en banc rehearing, the decision of which is here reported, acted otherwise than properly. 28 U.S.C. § 455 (“Any ... judge ... shall disqualify himself in any proceeding in which his impartiality might reasonably be questioned.”). Patently a judge who is disqualified from acting must not be able to affect the determination of any cause from which he is barred.5

The court en banc, therefore, could not have had among its members Judge Ervin. He, in recusing himself, has not infringed the statutory mandate that all circuit judges in regular active service shall comprise an en banc court. Any seeming contradiction between court action and statutory language is at most apparent; it is not real. There is a ready explanation. Judge Ervin is regular and active,6 and as a general proposition is in service. However, should he, or any other regular, active member of the court, recuse or disqualify himself at any time, he is out of service insofar as that particular case is concerned. To disqualify means to debar legally. See Webster’s New International Dictionary, 2d ed., p. 753. That is synonymous with lack of legal capacity, i.e., with inability to serve. Hence, Judge Ervin properly did not sit during the en banc rehearing, because, for the particular case, he was not one of the circuit judges in regular active service.

With that proposition established, we turn to the language describing the body, a majority of which is required to order hearing or rehearing en banc. In § 46(c), in a sentence immediately preceding the one stating who shall sit in an en banc hearing or rehearing, the body is described in substantially identical terminology: “the circuit judges of the circuit7 who are in regular active service.” Again Judge Ervin does not meet the description because, for purposes of the present case, he is “out of service.” It would obviously contradict the purpose of disqualification to treat the situation precisely as though the disqualified judge had voted “No.” The canon of construction is well established that words repeated within the same statutory section have an identical meaning in the several places employed. Eg., United States v. Nunez, 573 F.2d 769, 771 (2d Cir. 1978), cert. denied, 436 U.S. 930, 98 S.Ct. 2828, 56 L.Ed.2d 774 (1978).

We are not unmindful of authority to the contrary. Zahn v. International Paper Co., 469 F.2d 1033, 1041 (2d Cir. 1972), affirmed on other grounds, 414 U.S. 291, 94 S.Ct. 505, 38 L.Ed.2d 511 (1973).8 However, the opinions in that case, both majority opinions and a dissent, concentrate on the policy considerations and do not allude to the language of the second sentence in § 46(c) or to the problems of directly conflicting meanings assigned to the same language, appearing cheek by jowl in adjacent sentences in a single statute. It is perhaps possible that members of the Second Circuit concentrated their attention, reasonably enough, on FRAP 35(a). That rule, headed “When Hearing or Rehearing in Banc Will be Ordered,” restricts itself, so far as present purposes are concerned, to a setting forth in haec verba of the language of the first *905sentence of § 46(c). Consequently, the conflict between the conclusion reached in Zahn and the language of the second sentence of § 46(c) was not so readily apparent.

The result later reached in Zahn had been described as the correct one in En Banc Hearings in the Federal Courts of Appeals: Accommodating Institutional Responsibilities, 40 N.Y.U.L.Rev. 563 (Part I), 726 (Part II) (1965):

Section 46(c) vests the power to provide for en banc proceedings exclusively in the active circuit judges of the circuit. Even if an active judge is disqualified from participation in an en banc decision, he still should be able to vote on the question of whether to hear the case en banc. A vote to sit en banc involves considerations of effective judicial administration and does not constitute a vote on the litigant’s claim. Also, allowing the disqualified judge to vote on the en banc question has the further advantage of allowing the full membership to determine more accurately the likelihood of subsequent disagreement with the panel decision.

Id. at 736.

However, the logic is badly frayed inasmuch as it is assumed that a recused judge would vote on whether a case should be heard or reheard en banc even though disqualified from voting on the merits.9 The Fourth Circuit has simply not operated in that fashion,10 and we do not perceive how a disqualified judge could do so without infringing the rule that he should take no action which would possibly affect the outcome of a case. To count as a vote his non-vote might well be decisive as to whether the case would be reheard.

Better reasoned, we find, are two law review notes favoring a result contrary to that reached in Zahn: Comment, Federal Jurisdiction and Practice, 47 St. John’s L.Rev. 339, 348 (1972) (“Upon disqualification, the Chief Judge should have properly been accorded no weight; his vote should have been neutralized by reducing the count of the regular active bench to seven and, thus, the four judges who voted for en banc reconsideration of Zahn should have .carried the day.”); Comment, In Banc Procedures in the United States Courts of Appeals, 43 Fordham L.Rev. 401, 420 (1974).11

Since a disqualified judge under § 46(b) is explicitly recognized as unable to sit on a three member panel and since a disqualified judge manifestly, for the reasons heretofore given, cannot sit on an en banc hearing or rehearing, it would ascribe to Congress a petulant inconsistency, devoid of any apparent explanation, were we to hold that a judge, though disqualified, should be treated as though he were part of the quorum for the purposes of the vote on whether to hear or rehear the case en banc. The inconsistency aspect would also apply should we turn to general customary practice. See H. *906Robert, Robert’s Rules of Order 339 (Section 43) (S. Robert ed. 1970):

Majority Vote — the Basic Requirement [T]he basic requirement for approval of an action or choice by a deliberative assembly, except where a rule provides otherwise, is a majority vote. The word majority means “more than half”; and when the term majority vote is used without qualification — as in the case of the basic requirement — it means more than half of the votes cast by persons legally entitled to vote, excluding blanks or abstentions, at a regular or properly called meeting at which a quorum is present.

(Emphasis supplied.)

Accordingly, we have proceeded to rehear the case en banc. As a consequence, five court members (Chief Judge Winter, Judge Russell, Judge Widener, Judge Murnaghan and Judge Chapman) have concluded that there were errors mandating reversal for new trials as to the correct amount of compensatory damages of the judgments below in the Arnold and Mihalek accident cases.12 Judges Hall, Phillips, Sprouse and Bryan, dissenting, would affirm. Undisturbed since no en banc rehearing was sought, or because we did not regard them as meriting en banc reconsideration, however, were the aspects of the consolidated case as to which the panel opinion was unanimous: (a) the award of a new trial as to compensatory damages in the Weston accident case and (b) affirmance of the judgments entered in favor of the United States and the air traffic controllers on the claims for indemnification or contribution.

The rationale for the majority decision en banc is largely to be found in the panel dissenting opinion. See Arnold v. Eastern Air Lines, Inc., 681 F.2d 186, 206 (4th Cir. 1982). It would serve no purpose to repeat. However, it must be added that, although the panel dissent restricted itself to a description of perceived reversible error solely on the basis of improper conduct of counsel for the plaintiffs and the United States, we en banc, see substantial error in the joinder for trial or refusal to sever of the accident cases on the one hand and the indemnification or contribution actions on behalf of Aetna and other insurers of Eastern.

Of course, our judicial system could hardly function if appellate courts liberally indulged in second guessing of trial judges. The trial judges are on the scene, and properly are allowed extensive discretion in deciding such questions as whether the disclosure that a defendant is insured, while normally a grounds precluding consolidation or mandating severance, may be overborne by other considerations such as those relied on by the panel majority.13 Nevertheless, a trial must remain fair to both parties, and such considerations of convenience may not prevail where the inevitable consequence to another party is harmful and serious prejudice. Molever v. Levenson, 539 F.2d 996, 1003 (4th Cir. 1976), cert. denied, 429 U.S. 1024, 97 S.Ct. 643, 50 L.Ed.2d 625 (1976). Consolidation, or refusal to sever, where prejudice results under the facts and circumstances of the particular case, amounts to abuse of discretion, constituting reversible error. Dupont v. Southern Pacific Co., 366 F.2d 193, 196 (5th Cir. 1966), cert. denied, 386 U.S. 958, 87 S.Ct. 1027, 18 L.Ed.2d 106 (1967).

There is no need for us to announce a rule of universal applicability, imposing as an absolute necessity the separate trial of every case which is a candidate for consolidation if consolidation results in revelation to the jury of insurance coverage for a *907defendant.14 Nevertheless, looking at the facts and circumstances of the particular case, we are satisfied that knowledge of Eastern’s insurance coverage, and awareness of the approximate quantum of recovery for virtually every other crash victim, implanted in the minds of the jury resulted in prejudice, almost surely prejudice from the outset and certainly prejudice after the trial had wended its way to conclusion.

Had the cases not been consolidated, or after consolidation for discovery had they been severed for trial, there would have been no opportunity for counsel for one of the crash victims to make his uncouth observation about: “the 19 insurance companies that have been busy buying off the claims of the other people.” 681 F.2d at 211.

Looking alone at the size of the verdicts for compensatory damages in the three accident cases: Arnold ($3,027,500), Mihalek ($1,137,500) and Weston ($847,000), they cause the average for 70 other victims, $314,000, to pale by comparison. While peculiar characteristics of the three cases might afford an explanation, the other possibility simply cannot be eliminated that the jury, freed from any restraining concerns that it should not irresponsibly overstate damages and possibly unduly interfere with Eastern’s capability to continue to do business, contemplated the pockets — so deep as to seem virtually bottomless — of large insurance companies (several of whom, it was emphasized, were “foreign”) as a capacious source for benefactions rather than as only a spring supplying just compensation for injuries actually suffered.

In my panel dissent, I was “ready to assume, for the purposes of this case that, despite difficulties which customarily would dictate separate trials, consolidation of all the cases was, nevertheless, at least theoretically proper, in view of the great dislocations that otherwise were in store for the Western District of North Carolina.” 681 F.2d at 207. However, since consolidation “necessarily meant a disclosure to the jury that Eastern Airlines was insured, and that insurance carriers, rather than Eastern Airlines itself, would bear all or a substantial portion of any award to Arnold, Mihalek or Weston” (681 F.2d at 207), I emphasized that consequently there was “a heightened need to insist on scrupulous fairness in all other aspects of the trial of the consolidated cases.” 681 F.2d at 208. In the actual proof of the pudding, it has now become clear to me that the fairness required was not possible to attain. So I am entirely content to join with my four colleagues in a determination that the consolidation also amounted to reversible error.

That then presents the question of whether Aetna and the other insurers as parties to cases also held to have been improperly consolidated logically are likewise entitled to reversal and remand for a new trial. We are satisfied that the answer to be given is in the negative, for the consolidation prejudiced Eastern, but not the insurers. The fact of insurance had to be disclosed in the insurers’ cases, and the fact of indemnification of Eastern by its insurers would not prejudice the insurance companies.

REVERSED AND REMANDED FOR FURTHER PROCEEDINGS IN CONFORMITY WITH THE VIEWS HEREIN EXPRESSED.

concurring:

I agree with the majority that because of his disqualification Judge Ervin should not be counted in determining whether a majority of active circuit judges voted for rehearing this case en bane. In my view, however, the issue of whether rehearing en banc was proper in this instance can be resolved on a more narrow and equally compelling ground. On November 1, 1982, Judge Butzner took senior status before the order granting rehearing en banc was entered on December 23, 1982. It is true, as the majority opinion points out, that Judge Butzner was in regular active service when he voted on the suggestion for rehearing en banc. But, in my view, the fact that he was eligible at the time he voted on the suggestion is irrelevant. The critical time for determining a circuit judge’s eligibility to have his vote counted is when the order on the poll is entered. Thus, I conclude that Judge Butzner’s vote should not be counted, because when the order granting en banc rehearing was entered, he was no longer a circuit judge in regular active service.

I find support for this conclusion in the statute and rule governing en banc rehearings, as well as in the applicable case law. 28 U.S.C. § 46(c) reads in pertinent part as follows:

Cases and controversies shall be heard and determined by a court or panel of not more than three judges ... unless a hearing or rehearing before the court in banc is ordered by a majority of the circuit judges of the circuit who are in regular active service. (Emphasis added)

F.R.App.P. 35(a) further provides:

Determination of Causes by the Court in Banc

(a) When Hearing or Rehearing in Banc Will be Ordered. A majority of the circuit judges who are in regular active service may order that an appeal or other *913proceeding be heard or reheard by the court of appeals in banc. (Emphasis added)

The critical verb in both the statute and the rule governing en banc review is “order.” The power to order a rehearing en banc is clearly confined to those circuit judges who are in regular active service. Moody v. Albemarle Paper Company, 417 U.S. 622, 626, 94 S.Ct. 2513, 2515, 41 L.Ed.2d 358 (1974); Allen v. Johnson, 391 F.2d 527, 532 (5th Cir. 1968).

In United States v. American-Foreign Steamship Corp., 363 U.S. 685, 688-89, 80 S.Ct. 1336,1338-1339,4 L.Ed.2d 1491 (1960), the Supreme Court construed the provisions of 28 U.S.C. § 46(c) and concluded that the crucial time for ascertaining a judge’s eligibility to participate in an en banc decision was fixed by the statute’s use of the verb “determined.” Based on this construction, the Court held that a circuit judge who assumed senior status after the grant of a petition for rehearing en banc, but before the case was decided, was not eligible to participate in the decision.1

Applying this same rationale to the present issue, the validity of Judge Butzner’s vote can only be determined at the time the en banc rehearing was ordered. Because Judge Butzner was not in regular active service on December 23, 1982, when the order in this case was entered, I can only conclude that his vote on the suggestion for rehearing en banc may not be counted in determining whether the required majority was attained. Thus, in my view, without even reaching the question of Judge Ervin’s status, at the time the order was entered there were at most nine circuit judges in regular active service.2 Because five voted in favor of rehearing en banc, the requisite majority was achieved and rehearing en banc was properly authorized.

In all other respects, I agree with the majority’s reasoning on the en banc portion of its opinion. As for the merits, the majority opinion correctly states that I would affirm the district court. I agree with the result reached in the panel decision.

Circuit Judge, dissenting:

I respectfully dissent, both from the court’s decision to rehear this appeal en banc,1 and from its ensuing en banc decision on the merits.

The decision to rehear this appeal en banc was in the first place an improvident one tested by principles that should, but did not here, constrain the court’s invocation of that extraordinary procedure.

The decision on the merits, displacing a panel decision which, whatever its merits, was based upon a detailed and fully articulated record review of purely discretionary trial court rulings, is, by contrast, best seen as a blunt, though unadmitted, exercise of the court’s supervisory jurisdiction. That appellate power is invoked here to condemn and deter by the most drastic possible means misconduct of counsel which, though concededly egregious when considered in mass and in isolation, is never demonstrated on any principled basis to have been the likely cause of any actual prejudice to the only party favored by the court’s en banc decision. The predictable result of this apparently intuitive exercise of raw supervisory power is a correspondingly indiscriminate remand for a new trial that in this multifaceted case introduces a new set of *914injustices, wholly ignored by the court, affecting parties who have had no real opportunity to suggest them to the en banc court.

On both points more is involved here than the ultimate result in this legally unexceptional diversity case. Each decision trenches upon matters of ongoing institutional concern for which, by these decisions, unfortunate appellate review practices in this court are encouraged and regrettable legal precedent may have been established.

Because of the broader implications of the decision, expression of a dissenting viewpoint on both decisions is warranted. I take the decision to rehear en banc and the decision on the merits in that order.

I

The standards by which courts of appeals are to decide whether to rehear an appeal en banc are concededly not subject to precise formulation and wholly consistent application. The controlling rule, Fed.R. App.P. 35(a), implementing the undergirding statutory grant of power, 28 U.S.C. § 46(c), tells us only that the procedure is “not favored and ordinarily will not be ordered except (1) when consideration by the full court is necessary to secure or maintain uniformity of its decisions, or (2) when the proceeding involves a question of exceptional importance.” There is enough flexibility built into the very text of this rule — in the word “ordinarily” and in the open-ended expression of “exceptional importance”— that it could not be claimed that the rule itself either compels or excludes rehearing en banc in any case.

But of course, we know — as the introductory “not favored” admonition directly tells us — that the procedure is intended by Congress and the Supreme Court to be used sparingly and with careful discrimination, and that some other standard than a purely subjective, eye-of-the-beholder perception of “exceptional importance” should constrain us in the matter. The Supreme Court, which wisely has deferred as a matter of deliberate policy to the courts of appeals’ discretion in invoking this procedure, see Western Pacific Railroad Corp. v. Western Pacific Railroad, 345 U.S. 247, 269, 73 S.Ct. 656, 667, 97 L.Ed. 986 (1953) (Frankfurter, J., concurring), has, however, adjured caution and circumspection upon us. So we have been reminded by the Court that “[e]n banc courts are the exception, not the rule”; that they should be “convened only when extraordinaiy circumstances exist that call for authoritative consideration and decision by those charged with the administration and development of the law of the circuit,” United States v. American-Foreign Steamship Corp., 363 U.S. 685, 689, 80 S.Ct. 1336, 1339, 4 L.Ed.2d 1491 (1960); and that the policy of the en banc rehearing statute is “ ‘that the active circuit judges shall determine the major doctrinal trends of the future for their court,’ ” id. at 690, 80 S.Ct. at 1339 (emphasis added). See also Church of Scientology v. Foley, 640 F.2d 1335, 1338-41 (D.C.Cir. 1981) (en banc) (Robinson, C.J., dissenting).

Notwithstanding these wise admonitions suggesting the existence of objective criteria by which we should discipline ourselves in this matter, it remains the fact that we are constrained only by that discipline. And it cannot be gainsaid that in practical terms the standard for invoking the en banc rehearing procedure remains — to paraphrase Chief Justice Hughes on the Constitution — whatever a sufficient majority of active circuit judges in a particular case considers it to be. For this reason, it is not possible to insist that there are any cases which are improper “as a matter of law” for en banc rehearing, and of course I make no such claim for this or any case. See United States v. Lynch, 690 F.2d 213, 215 n. 22 (D.C.Cir. 1982) (Robinson, C.J.).

It is possible, however, to insist that there is a discernible scale of case-types in terms of probable fitness for en banc rehearing. We need periodically — if for no other reason than possibly to stem the tide of almost routine suggestions these days for rehearing en banc2 — to consider that scale as an *915aid to principled and sparing invocation of the procedure. This case provides a particularly appropriate occasion for the exercise because it involves the most flagrant disregard of the principles of restraint that the scale teaches.

Without attempting to define in detail the categories of cases most fit for en banc rehearing,3 it is certain that lowest in the order is that general category wherein the suggestion is only of non-constitutional error leading to an arguably unjust or incorrect panel decision having no significant precedential implications. As to such cases, former Chief Judge Kaufman has expressed the general view that a “judge should [not] cast a vote for reconsideration by the entire court merely because he disagrees with the result reached by the panel ... Rule 35 was not adopted to provide that luxury.” Gilliard v. Oswald, 557 F.2d 359, 359 (2d Cir. 1977). This bespeaks a general perception, with which I certainly agree, that it is not the intended function of the en banc procedure simply to provide in effect another intermediate appellate court to review for “mere” panel error. See Western Pacific Railroad, 345 U.S. at 275, 73 S.Ct. at 670 (Jackson, J., dissenting) (use of procedure for this purpose likened to “appeal from the three-judge court to a swollen circuit court”); Church of Scientology, 640 F.2d at 1341 (Robinson, C.J., dissenting); United States v. Robinson, 560 F.2d 507, 526 (2d Cir. 1977) (Feinberg, J., dissenting). This case is obviously in that general category.

Within this general low-fitness category of mere error-correction cases, there is concededly the potential in any case that the decision suggested for en banc reconsideration not only involves prejudicial panel error, but affects “doctrinal trends” or involves matters of significant public interest as well, so that “exceptional importance” justifying en banc rehearing on the latter basis is present. Even within this category of “correction-plus” cases, however, further gradations of potential “importance” can be identified and should be kept in mind.-

At the highest level of potential importance in the general low-fitness category must be cases involving misapprehensions or misapplications of significant legal principle, even of non-constitutional dimensions. These principles may in some cases be of such precedential moment that en banc rehearing is justified, not “merely” to correct an arguably erroneous panel result, but to bring to bear the full court’s consideration of a major doctrinal problem having serious precedential implications for circuit, and possibly national, law. See, eg., Adams v. Proctor & Gamble Manufacturing Co., 697 F.2d 582 (4th Cir. 1983) (en banc) (effect on individual Title VII claims of consent judgment in EEOC action). See generally Walters v. Moore-McCormack Lines, 312 F.2d 893, 894 (2d Cir. 1963) (en banc) (Lumbard, C.J.).

Certainly at a distinct level below such “important legal precedent” cases are those wherein the dispositive issues are purely factual or discretionary ones and it is suggested only that the panel incorrectly upheld or rejected trial court fact findings or found or failed to find abuse of trial court discretion. In neither type case is a legal issue with significant precedential implications likely to be involved.

In fact-dispositive cases, even if the controlling legal principles are of the greatest significance, rehearing en banc simply to consider a suggestion of panel error in its review of trial court findings of predicate facts is not warranted under the procedure.4 *916See, e.g., United States v. Collins, 462 F.2d 792, 802 (2d Cir. 1972) (en banc) (rehearing en banc improvidently granted where appeal turns solely upon validity of trial court factual determinations that controlled application of unchallenged Miranda principle).

At an even lower level — indeed by any reckoning at the very lowest level of potential importance — are those cases, such as the instant one, wherein the suggestion is only of panel error in affirming trial court discretionary rulings.5 Here the en banc court will simply be second-guessing a panel’s judgment call that was itself required to be greatly deferential to that first made at the trial court level. Such cases are not likely ever to provide a fit occasion for invoking the en banc procedure, no matter how egregiously “wrong” the panel’s judgment call affirming trial court discretion may appear even to a sizeable majority of the full court. No legal principle will emerge in clarified, altered, or corrected form from such an en banc decision.6 No general precedent of ongoing value in the exercises of trial court discretion or of appellate review of such exercises is likely to result, given the inevitably high fact-specificity of the trial court rulings.7 The only possible justification for rehearing such appeals en banc will be that found in the narrow, one-time-only “correction” of an “incorrect” or “unjust” panel decision affirming a trial court’s discretionary conduct of a particular trial whose exact dimensions will never be repeated.

Even if the possibility of justification in some discretion-review cases be admitted, the value of mere one-case “correction” in a particular case must always be weighed carefully in the balance against weighty conflicting considerations that underlie the basic admonition that the procedure is “not favored.” Those conflicting considerations are the obvious and simple ones identified by Justice Jackson in the course of pointing out why all rehearings en banc for merely one-case corrective purposes are disfavored:

[They] are not appropriate where the effect is simply to interpose another review by an enlarged Court of Appeals between decision by a conventional three-judge court and petition to [the Supreme] Court. Delay, cost, and uncertainty, which take their toll of both the successful and the unsuccessful, the just and the unjust litigant, are each increased by an additional appeal to a hybrid intermediate court.

Western Pacific Railroad, 345 U.S. at 273, 73 S.Ct. at 669 (Jackson, J., dissenting); see also Church of Scientology, 640 F.2d at 1341 (Robinson, C.J., dissenting) (“drains judicial resources [and] leads to multiplicity of opinions [resulting in] inability to offer authoritative guidance”).

*917With these considerations in mind, the en banc decision in this case confirms every point that makes rehearing en banc in this type case, of all types, most inappropriate. Laying aside the question whether the decision may have “corrected” a sufficiently grave panel error to justify on that basis alone invoking the generally disfavored procedure — a matter addressed in Part II — the decision predictably serves none of the dominant institutional purposes of the procedure.

It obviously resolves no intra-circuit conflict of decisions. It neither announces, abandons, nor clarifies any legal principle that might bear upon general “doctrinal trends” in circuit law. It lays down no precedent likely to provide helpful guidance either to the district courts in exercising discretion or to this court in reviewing exercises of discretion in future cases of general comparability. It deals with no controversy of such general public importance, as opposed to public curiosity, that full-court imprimatur is desirable simply to encourage greater public confidence in the end result. It simply substitutes the judgment of a one-vote majority of the nine-member en banc court for that of a one-vote majority of a three-member panel on purely judgmental questions of trial court discretion not likely ever to recur in even generally comparable form. Raw power obviously exists in the larger majority to do this; the real question is the propriety of the power’s exercise in this or other cases of the general type.

If one looks to the opinion itself — principally to the panel dissent which, it is said, contains the basic rationale for the en banc decision — it is obvious that the “exceptional importance” of the case was thought to lie only in the special venality of counsel’s misconduct, the special degree of dereliction in the trial judge’s failure to curb or correct it, and the special degree of panel error in failing to set things right at the appellate level. It could be asserted, I suppose, that on the root matter — counsel misconduct— the decision does embody a sufficiently important general principle and set a sufficiently important and helpful precedent. The principle, presumably: that counsel misconduct sufficiently egregious will taint and require setting aside a verdict; the precedent: that repetitions of this particular degree will in this circuit have that result, even if it takes en banc rehearing to achieve it.8

The general principle is of course unquestionable. But it has never been in doubt and convening an en banc court simply to emphasize its continued vitality is not a provident use of heavily pressed judicial resources.

Just as the legal principle is too general to justify two levels of appellate review to insure its “correct” application in a particular case, so the only legal precedent that could result from the effort will necessarily *918be so narrowly fact-specific as to be positively mischievous rather than helpful (if taken at all seriously). I predict on that score that the surest consequence of this decision will be a spate of appeals, petitions for rehearing, and suggestions for rehearing en banc in which we are invited to liken or contrast in minute, fine-honed detail various examples of counsel misconduct with that in this case. The result of that, I further predict, will be that so far as its precedential effect is concerned, this decision will quickly either be regretted or disregarded — equally unfortunate consequences for the stability and predictability of our processes.

There remains a possibility that, in all fairness, should be considered. Even conceding the general unfitness of this type case for en banc rehearing, and conceding that in a particular case no justification exists beyond one-case error correction, the injustice of a particular panel decision might be so grave and so manifest that a court armed with the power to set it right en banc could not responsibly fail to do so. I concede the point. See United States v. Lynch, 690 F.2d at 215 n. 22.

Where this is to be the sole basis for rehearing en banc, however, a special discipline is compelled for making the collegial decision that grave and manifest injustice rather than merely arguable panel error is involved. Fortunately, one reliable means of discipline is readily available. It lies in the very cast of the court’s vote on the en banc rehearing question. In a case where every member of a court other than the majority of a divided panel votes to rehear en banc, the sheer vote balance suggests a collegial judgment that injustice going beyond mere error is not merely arguable but manifest, and the validity of individual perceptions to that effect as reflected in the individual votes is by the overall vote confirmed. If, on the other hand, a vote to rehear en banc in such a case is close — particularly, as here, a bare majority — a reconsideration of the propriety of en banc rehearing is suggested for a collegial court by the very closeness of the vote. The strong indication from such a closely divided corporate judgment is that whatever of panel error is arguably involved, it does not amount to injustice so grave that its correction by the full court’s consideration is both necessary to insure justice and essentially foreordained by the vote.

In such a case, and I suggest in this case, a court should yet step back. It is never too late to determine, on the basis of newly revealed aspects of a case, that an original vote to rehear en banc was improvident when considered in a new light shed by later developments. See, e.g., United States v. Collins, 462 F.2d at 802-03; see also Church of Scientology, 640 F.2d at 1344-45 (Robinson, C.J., dissenting). This is classically such a case, and even at this stage of the proceedings, I think we should, upon reconsideration, deny the suggestion for rehearing en banc and let stand as the court’s judgment the panel decision. See id.

Aside from insuring in the end a more principled administration of the disfavored en banc procedure in this case, this would also yield the proper result on the merits. For, as I believe will be shown in Part II, the most that could be said for the en banc decision is that it simply substitutes for any injustice in the panel decision at least equal, and arguably greater, injustices running in the opposite direction. Where this is the only consequence of an en banc rehearing of trial court discretionary rulings, it confirms the improvidence of the initial decision to rehear en banc. It also points up the wisdom of the rule that accords great appellate deference to trial court discretion in balancing the risks of relative injustices (the very essence of discretionary judicial action) in an attempt to achieve the most just solution possible. In such situations, the questionable ability of appellate courts in general to strike any better discretionary balance cautions doubly against the expense and investment of resources involved in two-tiered efforts to do so. See generally, Wright, The Doubtful Omniscience of Appellate Courts, 41 Minn.L.Rev. 751 (1957).

*919This case should never have been wrung through this most dubious process to yield so dubious a substitution of pure judgment calls by so close a final call and with such new and unexplored injustices.

II

The en banc decision on the merits is, apparently, that largely because of the special risks of prejudice created by trial consolidation the misconduct of counsel took on such special color that the verdicts of all three plaintiffs against Eastern Airlines must be set aside and new trials on those claims ordered.9 The displaced panel decision rested on a contrary assessment that despite the conceded impropriety of much of the challenged conduct, a careful review of the record, undertaken with proper deference to trial court discretion, revealed that improper influence of the jury could not be assumed a probable consequence of the misconduct. Arnold v. Eastern Air Lines, 681 F.2d 186 (4th Cir. 1982) (Arnold I), superseded, Arnold v. Eastern Air Lines, 712 F.2d 899 (4th Cir. 1983) (en banc).

I stand on the panel majority’s assessment to that effect. Not only does it reach a more defensible result but, quite as significantly for our ongoing processes, it reflects a more principled process of review for this type case than is reflected in the en banc decision by the court. Because I think the en banc decision is flawed not only in its result but, just as critically, in its apparent process of review and correction, I add some thoughts here on the latter aspect.

A

The en banc decision is — so far as can be discerned from what is said in the two opinions — rested entirely upon a blunt perception that the misconduct of counsel here is so bad that review need not go beyond a simple recognition of its particular degree of impropriety — with remand for new trial dictated as a matter of course because of the degree perceived and the need to deter repetitions. No attempt is made to assess whether, under a careful record review, it must further be assumed that actual prejudice to Eastern probably resulted from the misconduct.10 I understand the court’s in*920stinctive first-blush dismay — even outrage — and have expressed my own.11 But I do not understand the court’s failure — as I read its opinion — to ponder more deeply the real issue that follows..

Review of trial court discretion in responding to party and counsel misconduct in jury trials is undoubtedly among the most difficult appellate tasks. See Arnold I, 681 F.2d at 194-95. The temptation in review is understandably great to make instinctive judgments that particular misconduct might improperly have influenced a jury because of its obvious capability to do so. But fair review requires a further effort on a principled basis to fathom whether it did in fact probably do so in the particular case.

The reason is obvious. We do not sit as general monitors of the professional conduct of lawyers in trial courts. In the overall scheme of things that important task is given over to others — to the organized profession in general and, in particular trials, to the broad discretion and equally broad final responsibility of trial judges. See Rosenberg, Judicial Discretion of the Trial Court, Viewed From Above, 22 Syracuse L.Rev. 635 (1971). Not only are we not given the task of general supervision, we are not equipped to do it fairly and consistently if we undertake it. Aside from other frailties and limitations that beset us as appellate judges in assessing the subtleties of both judicial and non-judicial influences on juries, the only effective sanction we have for correcting misconduct of counsel — ordering a new trial — is not available at all against losing counsel (surely equal culprits over the long haul and the run of cases) and, when invoked against prevailing counsel, is visited mainly upon their faultless clients.

That in the overall scheme it is not given to us to root out this particular evil whenever and however it comes to our attention in cases properly before us may be a matter of deep regret to different ones of us from time to time. But the temptation to range beyond our limited function of reviewing for judicial error should be resisted. The way systematically to resist it lies precisely in doing what I am satisfied was not done here: making pragmatic inquiry past the misconduct into whether improper influence upon the jury from that misconduct can be accepted as a probability, see City of Cleveland v. Peter Kiewit Sons’ Co., 624 F.2d 749, 756 (6th Cir. 1980), being content to leave to other forces of correction — or to oblivion— any misconduct not found legally prejudicial as well as ethically opprobrious. That exhausts both our responsibility and our power to correct for legal error.

It is critical that such a tough-minded and disciplined inquiry into actual prejudice was not — so far as we are given to understand — considered necessary here. For, as the majority panel opinion demonstrated, such a review rather easily dispels the initial shock effect undoubtedly created, as intended, by appellate counsel’s adroit massing, and heavy-handed repetition, of the various discrete instances of counsel misconduct upon which Eastern’s appeal was primarily based.

Full recapitulation of the panel majority’s review on that score is not appropriate here. It suffices to refer to some of the more obvious perceptions opened by that review and apparently considered inconsequential by the en banc majority: the significance of the extended period of time over which the instances of misconduct massed in appellant’s brief actually occurred, see Arnold I, 681 F.2d at 195,200; the numerous, carefully phrased cautionary instructions given by the trial judge designed to screen the jury from improper influences flowing from improper or questionable conduct by counsel *921for all parties, see id. at 198-99 & nn. 11-13; the fact that the dominant issue in the case was the degree, conceded to be at least “ordinary” negligence, of Eastern’s culpability, see id. at 197-99; and, most critically, the failure of Eastern’s counsel to have made any contemporaneous or other objection at trial to any but a very few12 of the numerous instances of conduct carefully massed for challenge on appeal, see id. at 200.

Because I remain persuaded that on a fair and properly deferential review of the record no prejudicial error resulting from counsel misconduct is shown, I would affirm the Arnold and Mihalek personal injury judgments and otherwise hold with the panel decision.

B

There remains a possibility that should be considered. It is that the en banc decision can be viewed and justified as a valid exercise of the court’s supervisory jurisdiction over proceedings in the district courts even if actual prejudice were not demonstrable. As indicated at the outset of this opinion I think this is the most forthright way to view the decision. Because I dissent from the result, however it may be grounded, I address this possibility.

Though the contours of this residual appellate power are understandably vague, it is undoubtedly drawn upon by federal courts from time to time in situations where justice, or the appearance of justice, has been thought to require corrective intervention even though actual prejudice to aggrieved parties cannot be demonstrated or is concededly absent. See, e.g., Thiel v. Southern Pacific Co., 328 U.S. 217, 66 S.Ct. 984, 90 L.Ed. 1181 (1946) (tort defendant’s verdict set aside for misconduct of jury commission and clerk of court in excluding daily wage earners from jury lists; “unnecessary to determine” actual prejudice to plaintiff); Young & Simon, Inc. v. Merritt Savings & Loan, 672 F.2d 401 (4th Cir. 1982) (new trial because of two-week interruption, over objection of counsel, of plaintiff’s presentation of complicated case; no assessment of actual prejudice). It is arguable that within this general usage of the power, a court may properly intervene to correct and deter sufficiently opprobrious misconduct by trial counsel — officers of the court — even though no actual prejudice stemming from the misconduct can fairly be demonstrated under ordinary standards for reviewing the trial court’s discretionary handling of the misconduct. Though the en banc majority does not say that this is the basis of its decision here, it seems to me implicit in what is said in the opinions giving the court’s formal rationale. See supra note 10.

Reserving the question whether the misconduct of private counsel for prevailing parties should ever be an occasion for invoking supervisory corrective powers — a point on which I have grave doubt13 — I am satisfied that in this case the misconduct does not. For two basic reasons.

In the first place, as developed in Part I, the existence of a sufficient degree of manifest injustice to justify an exercise of this *922extraordinary power is simply belied by the closely divided corporate judgment of the court on the en banc rehearing question and now, on the merits. Beyond this the remedy used to correct and deter the perceived injustice is itself afflicted with injustice in the overall perspective of this case.

The danger of giving an unjustly indiscriminate remedy is present in all exercises of supervisory jurisdiction. By definition, such exercises of judicial power are not focused upon the correction of specifically demonstrated prejudice to identified parties but upon the general regulation of conduct. That danger was present with a vengeance here and has been realized. A brief recapitulation of critical elements in the procedural development of the case illustrates this.

At the trial of these cases, the three plaintiffs had live, powerful claims for both compensatory and punitive damages arising from an almost incredible act of negligence on the part of Eastern’s pilots. So powerful were these claims and so overwhelming the available proof of negligence that — in what must surely be a relatively rare occurrence — this major commercial airline had conceded ordinary negligence entitling the plaintiffs to compensatory damages, leaving for trial on their claims only the amount of compensatory damages and the existence and amount of any punitive liability. Following three weeks of trial and seventeen hours of deliberation, a jury gave substantial compensatory awards to the plaintiffs, but denied them punitive damages.

The effect of the en banc decision is not only to strip these plaintiffs of their compensatory damage awards, but to send them back stripped as well of any opportunity to establish before the jury that reconsiders their compensatory awards their possible entitlement as well to punitive damages. This misfortune cannot be chalked up simply to the normal hazards of litigation and the inscrutability of jury processes. It stems rather from a circumstance that calls for relief if retrial of all the compensatory awards is to be the result of this appeal.

The compensatory awards are stripped because of the misconduct of counsel for these plaintiffs. But the loss of opportunity to attempt new proof of punitive liability results from a bizarre procedural twist in post-verdict proceedings when both Eastern’s motion for new trial on the compensatory awards and the plaintiffs’ motions for new trial on the punitive damage issue were being considered. That twist is not fairly chargeable either to plaintiffs or to their counsel and it may well have led to the crudest injustice experienced so far by any parties to this action — in this matter, the plaintiffs. It came about in this way.

Despite the trial judge’s finding, as trier-of-fact on a related, essentially congruent, non-jury issue, that “[Eastern’s] crew were not simply inadvertent but rather were grossly negligent,” see Arnold I, 681 F.2d at 190 n. 1, 194 n. 8, the judge declined to set aside as against the weight of the same evidence the jury’s verdict against plaintiffs on their claims for punitive damages based on the alleged gross negligence of that crew.

Though it is, of course, difficult to be certain of the reasons for this apparent anomaly in the judge’s rulings, it seems almost certain that they represent a practical assessment that the high-side compensatory awards on the one hand and the rejection of any punitive damage awards on the other reflected a jury “compromise.” This is indicated both by the judge’s reference to the possibility, see id. at 204 n. 16, and by the next sad twist in the ironic spiral into procedural disaster now underway for these completely faultless plaintiffs. Of course, if the assessment was of compromise verdicts, the proper technical response was to set both aside and order new trials on both compensatory and punitive damages. But in what now appears an effort to solidify the jury “compromise” as essentially a just one, the trial judge next admonished Eastern’s attorneys on the record that if they successfully appealed the compensatory awards they would face on any new trial they won not only the compensatory damage claims but the punitive damage claims as well. See id. at 206 n. 22. Although conditionally ordering such a consequence *923while declining to set aside the verdict did not lie within the judge’s power, see id., it is understandable that counsel for plaintiffs may reasonably and responsibly have assumed that he did have the power. Presumably acting on that assumption, counsel failed to take a cross-appeal, and the unfolding of plaintiffs’ procedural misfortunes continued. Technically, of course, the absence of a cross-appeal precluded review of the judge’s refusal to set aside the verdict denying punitive damages, and the panel so noted in affirming the two personal injury awards and remanding the Weston wrongful death claim. See id.

That technical impediment to cross-relief still exists of course. But it would be a singularly unjust withholding of the court’s supervisory power not to grant that cross-relief now as an adjunct to the substantial relief given Eastern in the process of eradicating the perceived injustice stemming from counsel’s misconduct. The improvidence of the trial judge’s ruling, with its inevitable lulling effect on plaintiffs’ counsel, should not be charged to those plaintiffs even if the misconduct of their counsel is now to be. Ample precedent exists for giving this relief, notwithstanding the technical barrier, in an exercise of the same supervisory jurisdiction that is arguably the most defensible basis of the court’s decision to benefit Eastern so handsomely for the dereliction of opposing counsel. See, e.g., Tug Raven v. Trexler, 419 F.2d 536, 548 (4th Cir. 1969); National Association of Broadcasters v. FCC, 554 F.2d 1118,1127 & n. 24 (D.C.Cir. 1976).

For this reason, I would at the least, if the compensatory damage awards are to be set aside, exercise this court’s residual power to make a more just overall disposition in this tangled web of procedural misfortune. As a part of the remand, I would direct the trial judge to reconsider, on the record already made, the plaintiffs’ motion to set aside the jury verdict finding Eastern not liable for punitive damages. The judge’s initial ruling on that motion was so obviously skewed by understandable concern to salvage verdicts possibly reflecting compromise, and his follow-up attempt to nail down the compromise so obviously misled counsel to plaintiffs’ detriment, that this result seems to me compelled by the same considerations of justice that the en banc majority has concluded require setting aside the compensatory awards.

A final injustice involved in the remand for a new trial so handsomely and solely to Eastern’s benefit should be noted. It further illustrates the improvidence of the specific decision to rehear en banc in these consolidated cases. And it underscores the unevenness that is likely to result from exercises of supervisory jurisdiction in multi-party, multi-claim cases where the effects of corrective action on all affected parties either are not, or cannot, be taken into account in shaping the remand.

This concerns the ironic plight of Eastern’s insurers (Aetna) in the final denouement. In their consolidated appeals Aetna heavily relied, as did Eastern, on the prejudice allegedly caused it by misconduct of opposing counsel. See Arnold I, 681 F.2d at 192-94. Aetna’s challenge on this score was rejected by the panel, as was Eastern’s, but in the case of Aetna, the rejection was by unanimous decision of the panel. On Aetna’s and Eastern’s ensuing petitions for rehearing and suggestions for rehearing en banc, Eastern’s suggestion for rehearing en banc of course carried by the narrowest of margins. At the same time, Aetna’s petition for rehearing was being denied unanimously by the panel, and its suggestion for rehearing en banc failed when it prompted no request for consideration by any member of the court. Now Eastern, because of counsel misconduct, has secured a new trial limited to the determination of compensatory damages and has been freed of the potential for punitive liability, while Aetna’s appeal based primarily on the same ground has ended in apparent failure, with the Supreme Court’s recent denial of Aetna’s petition for certiorari following our denial of its petition for rehearing. Aetna Casualty and Surety Co. v. United States, cert. denied,-U.S.-, 103 S.Ct. 1801, 76 L.Ed.2d 366 (1983).

*924It is arguable, and indeed I have always believed, see Arnold I, 681 F.2d at 204, that Aetna’s claim of prejudice from counsel misconduct was even less supportable on a careful record review than was Eastern’s. But that was on the basis that, as Judge Bryan and I believed, neither was supportable. It is frankly inconceivable to me — and I am sure to Aetna — that if Eastern’s claim to relief on that score has sufficient merit to justify en banc corrective action, Aetna’s had not enough to merit the same.14 The record simply will not support the drawing of that fine a line between the relative positions of these parties on that matter.

On the overall record of these proceedings, Aetna could be excused if it wondered whether somehow its claim for appellate relief simply fell through the cracks of the en banc procedures in this court.15 And any fair legal analyst would have to conclude that either this court’s decision to rehear Eastern’s appeal en banc or its failure to rehear Aetna’s based upon the same ground was improvident. I have, of course, already indicated which of these I think was improvident. But one surely must have been.

Whatever its other consequences, this case reveals the special perils of indiscriminate decisions to rehear en banc and consequent indiscriminate remands in multiclaim, multi-party cases of this procedural complexity. The trial court’s discretionary resolution of the admitted dilemmas created by trial developments remains the more defensible one, as the panel long ago concluded.

I am authorized to say that Judge SPROUSE joins in this opinion.